Experiments with knockouts of the genes encoding the D-type cyclins demonstrate that breast cancers induced by the neu or ras oncogenes in mice are absolutely dependent on cyclin Dl. Most human breast cancers overexpress cyclin Dl: approximately 50% of human breast cancers have amplified or otherwise overexpress c-neu, which depends on cyclin Dl for cell-cycle activation, and approximately 20% of human breast cancers have amplified cyclin Dl. A logical approach to the treatment of neu-, ras-, and/or cyclin ZX?-associated breast cancers would be to inhibit cyclin Dl function. Because knocking out cyclin Dl has a limited impact on adult mouse physiology, specific inhibition of cyclin Dl would likely have a very favorable therapeutic index. Cyclin Dl functions by direct interaction with other proteins. Protein-protein interactions have been inhibited successfully using small molecules identified in library screens. To maximize therapeutic index, screens to identify specific inhibitors of cyclin Dl interactions should be designed in such a way as to exclude inhibitors of cyclin D2 and cyclin D3 interactions. Given the sequence divergence among the D-type cyclins, it should be possible, using reverse two-hybrid screening in yeast, to identify inhibitors of cyclin Dl interactions that do not inhibit cyclin D2 and cyclin D3 interactions. The screens in this proposal are designed to select for inhibitors of cyclin Dl interactions by linking these interactions to the expression of genes conditionally toxic to yeast. To exclude inhibitors of cyclin D2 and cyclin D3, the interactions of these cyclins will be linked to the expression of genes conditionally required for yeast to grow. Yeast have been used successfully in drug screens, as they are permeable to compounds in chemical libraries, which average less than 500 Da. Reverse two-hybrid screens are straightforward to construct and validate, and can be adapted to high-throughput screening of small-molecule libraries in 96-well plates, or to screening peptide-aptamer libraries on solid media. Specific inhibitors of cyclin Dl could be tested further in mammalian-cell-culture and mouse models, and would be useful both as research tools and as potential therapeutic agents for neu-, ras-, and/or cyclin Dl -associated breast cancers. [unreadable] [unreadable]